So, I have to say, Vivek, when you asked me to do Diagnostic EOS, I was really scratched my head, like, how am I going to do this in 20 minutes? Because you know, it takes a year to learn this stuff. So I've really like, again, focused on some take home points, instead of doing a really all encompassing presentation. So again, I don't have any disclosures. I'm going to focus on some general tips. I'll talk about some mainstream uses. We'll go through a couple of more common uses of Diagnostic EOS. And then I'll concurrently discuss some challenges with each of these. So when we talk about general tips, there are a few of these. The first one is know your equipment. And in the case of EOS, I think it's very essential that you have a sense of the spatial relationship of the scanning range to the anatomy. And obviously, this is different for radial scopes and for linear scopes. But it's very important that you have that sense. Because when you do endosynography, and you're looking at images tangentially, you can overestimate or underestimate things. So have that sense of relationship as you are doing in EOS. The second tip is that, you know, that we have a keyboard, right? And there's a lot of stuff on this. I'll be very honest, I don't even use half of it, I probably don't even know what half of it is. But there are three things that we use the Doppler, which most people know, it's really a yes and no. Is there a vessel or isn't there a vessel? And I don't really get into too many more details, others might. But the two that I use more commonly are the gain and frequency. So the gain button is right over here, it essentially tells you whether the image, it essentially adjusts the brightness of the image. As you can see over here, this is the kidney. And this the gain is very high. So you can see that the image is very bright, but it's also very noisy. And on this side, the same image has a very low gain. So the image is much darker, and you cannot see the margins as well. The only reason I adjust the gain is typically when I want to define the margin and its relationship to the surrounding tissue. It comes with a default gain of around 12 to 13 on the, and I usually leave it as is, it's only when I'm trying to really finagle the details that I play around with it. The second one is frequency. Now this I use a lot, and I really recommend that anyone who's doing endoscenography should be very familiar with this. Frequency basically tells you the degree of penetration of the waves into the tissue. And it's inversely proportional to the resolution. So the higher the frequency, the higher the frequency, the lower the penetration and the higher the resolution. So I'll give you an example. So here we have different frequencies. So you can see at 12 megahertz, the resolution is much higher, but your depth of penetration is really not that deep. At seven and a half, which is sort of where we live when we're doing our scanning, the penetration increases, but the resolution drops. And then at five, you can see, you can look pretty deep down into the extramural area, but the resolution is really gone. So the way I use it is if I want to look at extramural tissues, then I use the low frequencies. When I'm doing my general scanning, I stay in the middle somewhere, seven and a half is usually where I live. And when I want to look at intramural lesions, then I typically increase the frequency up to 12, because I really want to delineate the different layers that we have. So I think it's well worth our time to familiarize ourselves with what frequency can do for us. The second, the third tip is really know your layers. I know this is sort of, this is the basis of a lot of the stuff that we do in EUS, but just as a quick reminder, there are five layers. I know that there's some literature out there about nine layers, but I won't get into that. That's typically for intramucosal esophageal adenoparcinomas, but for all intents and purposes, the five layers, which are the white, dark, white, dark, white, but in more scientific speak are deep mucosa, muscularis mucosa, the muscularis submucosa, muscularis propria, and then the extramural tissue. So you really need to know this when you do intramural lesions. And then there's some really, you know, basic tips, which is have a home base in whichever organ that, you know, you are examining. It's typically the portal vein in the duodenum, the celiac axis from the stomach, aortic arch and the esophagus. Every time you're lost, just go back to the home base and start again. Pay a lot of attention to the normal. I know when we train fellows, we focus a lot on what abnormal looks like, but to be honest, abnormal can sometimes be extremely subtle. So if you don't know normal, you will not be able to tell abnormal. So I say pay a lot of attention to what normal looks like. Also remember that every patient is different and the same lesion can look very different in different patients. And I'll tell you, neuroendocrines and carcinoids are really, they can really flummox us even to this day. Air is your enemy, water is your friend. Basic truth in EUS, every time you meet resistance, just stop. Every time I've seen a perforation in EUS, it was because there was some kind of resistance and you know, the EUS was not stopped. The most common areas where you'll meet a resistance is the UES and the D1, D2. And especially in the cases of tumors, D1, D2, any resistance, just stop. I actually have my fellows use the ERCP shortening maneuver right to the D1, D2 turn, rather than pushing. And then finally, I think this is very, very important for diagnostic EUS is that it's not a fishing expedition. You know, you will get these questions from your colleagues, hey, can you take a look down and see what's going on? Absolutely not. EUS is really, you have to have a pretest probability of what the EUS is going to show you. You have to have a crystal clear question in your mind of what you hope to answer. And then you basically come up with a clear answer of what you found. So it really should not be used as a fishing expedition. And I'm telling you that these are some of the requests that we do get, especially in cases of chronic abdominal pain. Now the current uses for endoscopic ultrasound are diagnostic, and I'm going to go over these. Girish is going to do a talk on pancreatic cysts, so I won't talk about that. And then imaging of tumors, tissue acquisition, and then interventional EUS. So I'm going to focus on the more common diagnostic indications. So the first one is submucosal lesions. So you know, you can see a variety of these, as you can see over here, that, you know, they can basically look like anything. And even after years of experience, I can tell you that we are sometimes wrong looking at the endoscopic appearance and trying to determine what it can be. So the questions you have is, what is it? Is it worrisome? Does it need surgery? Should we ignore it or should it be surveyed? Out of all these four pictures, you know, I can make my guesses and I'm sure most of you can too. The only one that I think I can say with confidence is this one, which is a pancreatic wrist. I typically in my practice don't recommend that this should undergo an EUS. The rest, you know, we can be wrong. So let's walk through these. Now, when you are looking at submucosal lesions, it really is a game of layers. You really have to have an understanding of the layers in order to be able to make a diagnosis. And now this is a kind of a busy graphic, but if you stay with me, these are the layers and this is the echo levels. So you have anechoic lesions that have no echo features at all. You have hypoechoic lesions, which are very, very dark. And then the hyperechoic lesions, which are very, very bright. All the different intramural lesions live within different layers and, but it's not in real life. It's not as simple as this. So if you have, for instance, varices, they will be anechoic and they typically live in the second layer or in the, in the deep mucosa similarly lipomas, which are very hyperechoic will live in the, in the third layer, which is the submucosa and gists again, which are very common and will live on the fourth layer. But sometimes you can not differentiate between a gist and a leiomyoma. So these can be actually quite tricky. I think the ones that you can really diagnose with confidence are the varices, because obviously you have Doppler and they will be anechoic lesions cysts again, because they will be anechoic, but sometimes they can look solid. And I'll, I'll discuss that in one of the cases. And then obviously lipomas, which have a very typical appearance because they have fat and fat is very bright. And when you see that you can be fairly confident that that's a lipoma. The rest can be a little bit trickier. So let's look at these cases. This was the first case that was sent to us again, incidental finding on somebody who underwent an endoscopy for GERD. This scar is actually from a previous biopsy, which understandably was unremarkable. We did an EUS and the EUS shows I have three arrows here. So as to not confuse it with the aorta, but it demonstrated this dark hypoechoic lesion coming off the last layer, which is the muscularis propria. So, you know, you can say with confidence that this is a gist. Now the question obviously is management. What are you going to do with these small gists that are, you know, less than two centimeters look very round and well, you know, round and regular well-defined margins and do not have any areas of cystic degeneration. These are typically considered the low risk gists. We enter them into surveillance protocols, but there's really no data out there that shows how they will behave or what the surveillance intervals should be. I know that some centers recommend just removing them with a full thickness resection. We typically don't do that at least at our institution. Now when you see another gist that looks like this with these areas of cystic degeneration that I have, again, with very, very irregular margins and sort of a heterogeneity to the pattern, that is indicative to me that this is a higher risk gist. And these gists we recommend be removed, even if they are less than two centimeters in size. Now, if I had seen this on endoscopic view and I'd probed this and I'd felt a very, very firm submucosal lesion, I could have made this deduction that this is most likely a gist, but you definitely need an EUS because the other like carcinoids can look like this as well. Now, the second one is really like a small kind of a submucosal lesion that was sent our way. And when we did the EUS, it actually is coming from the submucosa. You can see this very dark layer, which is completely intact behind it, which is the submucosa. And this dark, which is hypoechoic, and we effinate this and this turned out to be a carcinoid. But you can see, I mean, you can imagine that if it was closer to the MPE, how you would have a hard time differentiating this between a gist. So this was a carcinoid. Then this lesion was, you know, this patient actually had signs of GI obstruction, gastric outlet obstruction, and this is what we found. We did an EUS on it. And again, you can see there's a very nice intact muscularis propria behind it. This lesion was coming out from the submucosa, has a very homogeneous pattern is quite bright. So this was a lipoma. And once you diagnose them with certainty, you can actually just disregard them unless they're causing obstruction, which was this patient's case in which we went ahead and did an EMR and remove this. This final one is somebody I saw recently actually with a submucosal lesion that was found again on a routine endoscopy. This was in the esophagus. And when we did the EUS, you can see that there's a complete wall layer pattern outside it. And this is essentially anechoic with some debris inside it. So this was most consistent with the duplication cyst. Now duplication cysts, once you're diagnosed them, you can leave them alone. There are some reports of malignant transformation, but for all intents and purposes, you don't have to do anything. The one thing I will say is that some of them can actually look solid. You see this debris over here, sometimes it can extend in the entire cyst and make it look very, very solid. And you know, we have been fooled. The one thing I will say is if you're going to biopsy a duplication cyst, especially in the esophagus, make sure you have antibiotics on board because you can actually cause severe mediastinitis. We had one such case here where the patient ended up having a very prolonged hospital stay and required mediastinum exploration. And that happened even with antibiotics on board. So that's something just to be aware of with these cysts. And this obviously, everybody can tell that these are varices, but this is just to show you that they have a very serpiginous, anechoic appearance and you'll see flow in them. Sometimes you will get these referrals when you have an isolated varice and the person is not sure whether this is something that is a mass or a varice. Now how accurate are we at telling that these submucosal lesions are what we think they are? Well, there's a study in which they looked at 100 patients and they found that we were accurate in only 48% of the cases. So we're actually not that good at telling, and most of the misclassifications happened in the hypoechoic lesions in the third and fourth layers, which is the neuroendocrines, the gist, the carcinoids, the ectopic pancreas, and granular cell tumors. And when we look at experts, so this is a compilation of inter-observer variability between experts for various lesions, and you want to have a kappa statistic or agreement, which is at least moderate. Ideally you want to be at above 0.6, which is good, and 0.8, which is very good. But as you can see that the best agreement was for cysts and for lipomas, and for everything else pretty much, it was only fair to moderate. So even the experts don't really agree, and this is why I always say that diagnostic EUS can sometimes be harder than interventional EUS. And then that is when you obviously can deploy EUS FNBs, EUS FNA and FNBs, and Vivek is going to talk about that, but that can also be tricky in these subepithelial lesions depending on the size. So the take-home point I think for EUS for subepithelial lesions is that it's helpful, but sometimes it needs to be complemented with a biopsy in order to make a definitive diagnosis. Now the next, I call this the champions in EUS, is chronic pancreatitis. Now these are, it's almost sometimes what feels like the wild, wild west in real life, but because you get all kinds of questions about diagnosing chronic pancreatitis. So the typical indications are that, you know, there is a clinical suspicion for chronic pancreatitis. The patient has had imaging which demonstrates it. You might see a patient who has acute recurrent pancreatitis and you're asked to rule out chronic pancreatitis. You will get patients who have chronic abdominal pain and, you know, the physician is really at a loss as to what this could be. And they want you to rule out chronic pancreatitis. And my favorite right now is chronic diarrhea because everybody now seems to be getting a fecal elastase on chronic diarrhea. And I'm spending all my time educating people that this is really not a standard algorithm, but there we have it. And they have a low fecal elastase and you're asked to rule out chronic pancreatitis. So in 1998, there was a set of standard criteria that were developed that would help us diagnose chronic pancreatitis. And these included parenchymal abnormalities and ductal abnormalities. And basically these were bright strands or bright foci, how lobular the pancreas was, looking at the main duct dilation, irregularity and the margins of the duct, and obviously looking for calcifications, which is intraductal stones or parenchymal calcifications. So this is an ideal picture of what chronic pancreatitis would look like. You have hyperechoic strands, as you can see these linear structures. You have hyperechoic foci, which is spots, essentially. You have lobulation of the pancreas. You have a slightly dilated duct with very bright margins, but that's not how it is in real life because you will have the same findings in older patients, in smokers, people who drink alcohols socially, diabetics, and it's very, very operator dependent. You can essentially make anything look like anything in the U.S. And I'm serious. I mean, you can make up stuff. So it's really, really important to be very aware of what you're looking for. So in 2009, the Rosemont criteria was developed for chronic pancreatitis, and the impetus behind this was to give a weight to each of these features, and then depending on a combination of weights, have sort of a stratification where there's a definite diagnosis, there's a suggested diagnosis, undetermined diagnosis, and then normal pancreas. So how good are we at agreeing amongst ourselves that these criteria are present or absent? Turns out not very good. So these are studies looking at both standard scoring as well as Rosemont scoring, and you can tell this is just fair to moderate agreement between experts. And I can tell you when the Rosemont criteria came about, we ended up actually printing it out and sticking it up in the room because it was so hard to remember. But again, we really can't agree amongst ourselves as well. And even when we compare these criteria to histology in patients with chronic pancreatitis, you can see that there's only weak agreement in most studies. And I'll show you some images to demonstrate to you how difficult this can be in real life, and this is why it's so important that you sort of have that pre-test probability in your head. So this is, you can see a gorgeous looking pancreas, homogeneous, really with nothing going on. So this is a normal pancreas. Now, if I saw this, I would have said this is chronic pancreatitis, right? Has foci, has some strands that you can see, dark, slight lobularity. This was actually not chronic pancreatitis. This was a smoker without chronic pancreatitis. Did he have early chronic pancreatitis? I don't know. Now, this patient, again, you would have said same thing. You see these hyperechoic reflectors, some lobularities, these dark spots. This was actually a diabetic patient without chronic pancreatitis. This patient had chronic pancreatitis because you can see a stone in a dilated, irregular pancreatic duct. So when you see this, you can be very confident that this patient has chronic pancreatitis. This patient, you can see the pancreas has a lot of lobularity over here with some hyperechoic stranding. You would be inclined to think this could be pancreatitis, or this is definitely not a normal pancreas. This was actually a patient who was BRCA positive and was undergoing screening for pancreatic cancer. No indication, no history, no symptoms, nothing for chronic pancreatitis. And finally, this patient had very, very normal-looking pancreas, as you can see parts of it here, but had pancreatic calcifications in the pancreatic duct. So this was somebody with chronic pancreatitis. So I think you get a flavor for how difficult this can be, and this is why I keep on going back that this is really not a fishing expedition, because you can sometimes be over-diagnosing chronic pancreatitis in patients. So I think the take-home point for chronic pancreatitis in EUS is this, for non-calcific chronic pancreatitis, the role is still evolving. We don't know what are the minimum criteria for a firm-based diagnosis of chronic pancreatitis. We think that some criteria have more weight, even if the pancreas looks normal, but I see class calcifications, I'll say this patient has chronic pancreatitis, and I think most of us would be right if we'd say that, but for all the others, it's really a constellation of clinical criteria that you need to put together before you make a diagnosis. We don't know if we can diagnose early chronic pancreatitis by EUS, we don't even know what the gold standard is. And also, like I said, all criteria may not be equally important, except calcifications, I think. So you may want to, again, take the whole clinical picture in mind. In clinical practice, what helps us the most is extremes, either a normal pancreas or an entirely abnormal pancreas. It's the in-between that actually really can be, you know, angst-provoking. So moving on to the role of EUS in biliary disease. This is somebody that I just saw actually last week. So this is a 53-year-old female with intermittent epigastric pain for five months. She got a CAT scan in the ER which showed polylithiasis and a CBD measuring seven millimeters. They did labs, ASD-ALT was mildly elevated, which normalized the next day. MRCP shows a CBD of seven millimeter without polylithiasis, and you get a call from surgeons who want an ERCP for clearance. So what do you do? So the ASG actually has criteria that helps us define or stratify or give a risk assessment to these patients. If the patient comes in with a CBD stone on imaging, you know, game over, right? It's a stone, you go ahead and do an ERCP in these patients. If they have acute cholangitis, ERCP. If they have a serum bilirubin greater than four and a dilated CBD on ultrasound greater than six, or without an ultrasound greater than eight, that is considered a high risk indication for greater than 50% probability of the patient having a stone. So in these cases, nobody's going to fault you for going straight to an ERCP. The second one is intermediate risk, which is you have abnormal liver chemistries, age greater than 55, and a dilated CBD on imaging. In these, if you have any one of these criteria, there's a 10 to 50% probability of the patient having a common bile duct stone. And this is where EUS lives. This is where you will step in and do an EUS with plans to flip to an ERCP in the same setting ideally to evaluate if the patient has a stone or not. And obviously if you don't have any of these predictors, the patient is at low risk. So looking at our case, you can see we did an EUS and there is a stone right in the common This is the portal vein, the common bile duct with the stone and shadowing right behind it. So we just immediately flipped it to an ERCP and did an ERCP. And then how good are we compared to cross-sectional imaging? This is a meta-analysis, which basically shows that our performance characteristics are better than an MRCP. And this is basically because the MRCP misses stones that are less than six millimeter in size. And that's why EUS can actually pick up smaller stones that an MRCP may miss. So here's, I've added this case. This is actually one of my cases from last year. That's interesting. This is a 74 year old male with a mildly dilated bile duct xenon imaging. He has a gallbladder. He has severe NASH with persistently elevated LFTs. And they've been repeating it. He was followed in the liver clinic and they were repeating MRI, MRCPs over two years. They were all negative except for the dilated CBD. And I was literally curbsided on this patient. And I said, you know, let's look at, do an EUS and see what's going on. So we did an EUS and lo and behold, we found this lesion in the bile duct. You can see there's no shadowing behind this. This is actually a mass. And I went ahead, did an ERCP and brushed it. This patient had intraductal adenocarcinoma of the bile duct. So sometimes in these kinds of iffy cases where there's a dilated duct and you really don't have a clear picture, you will have meaningful information with an EUS, either a small ampullary mass or something growing inside the common bile duct. So again, another area where EUS can be very, diagnostic EUS can be very helpful. The complications overall, the number that I quote is very similar to a regular endoscopy, less than 1% overall. Perforation is the scariest because it has mortality associated with it. Again, like I said, in one of the earlier slides, if you meet resistance, don't push. The two main sites are the UES and the fixed duodenum, especially in the cases of tumors. There are obviously complications associated with EUS FNA, zero to 3%. I almost never seen bleeding, knock on wood, after an EUS FNA, but I have seen pancreatitis. And I think Vivek is going to talk more about EUS FNA and FNB. So in conclusion, you know, EUS definitely has a role in the diagnosis of various lesions. It has a definitive role in the diagnosis of lipoma, cysts, and varices. It has a complimentary role in other submucosal lesions. Chronic pancreatitis, you have to really pick the right patients for the right reasons. Look at all the clinical signs and symptoms to come up with, like I said, a pretest probability of what you're going to find. The extremes are most helpful, a normal or an abnormal pancreas will help you. The in-between are not that helpful. And I don't think we should really be using EUS to make a definitive diagnosis of early chronic pancreatitis, as I demonstrated in the other slides. There's a definite role of diagnostic EUS in patients who are at intermediate risk of choledocolithiasis, and then in selected patients with dilated common bile ducts. You will get literally, there's literally, the doors will open once you start doing this. Anybody with a dilated duct will be getting an EUS, but you have to be really careful about what the reason is for doing this EUS. And again, finally, I always say this, it's not a fishing expedition. So thank you very much, and I hope I'm on time Vivek.

diagnostic endoscopic ultrasound

equipment knowledge

gain button

frequency button

scanning range

anatomy